{"passport":{"unfragile":{"@version":"1.0","version":"2026-05","artifact":{"id":"npm_npm-codex-mcp-server","slug":"npm-codex-mcp-server","name":"codex-mcp-server","type":"mcp","url":"https://www.npmjs.com/package/codex-mcp-server","page_url":"https://unfragile.ai/npm-codex-mcp-server","categories":["mcp-servers"],"tags":["mcp","codex","openai","claude","ai","cli"],"pricing":{"model":"open_source","free":true,"starting_price":null},"status":"active","verified":false},"capabilities":[{"id":"npm_npm-codex-mcp-server__cap_0","uri":"capability://tool.use.integration.mcp.wrapped.codex.cli.invocation.with.streaming.output","name":"mcp-wrapped codex cli invocation with streaming output","description":"Wraps OpenAI's Codex CLI tool as an MCP server resource, translating MCP protocol calls into local CLI invocations and streaming results back through the MCP transport layer. Uses child process spawning to execute Codex commands with environment variable injection for API credentials, capturing stdout/stderr and marshaling responses into MCP-compatible JSON structures for consumption by MCP clients like Claude.","intents":["I want to use Codex code generation capabilities from within Claude or another MCP client without direct API calls","I need to expose a local Codex CLI installation as a standardized MCP resource for multi-tool orchestration","I want to integrate Codex into an agentic workflow that already uses MCP for tool discovery and invocation"],"best_for":["developers building MCP-based agent systems who want Codex as a pluggable code generation tool","teams running Claude with MCP servers and needing local code generation without cloud API dependencies","builders prototyping multi-LLM workflows where Codex specialization is valuable alongside Claude reasoning"],"limitations":["Requires local Codex CLI installation and valid OpenAI API key — no fallback to cloud-only execution","Streaming output depends on Codex CLI's own streaming capabilities; latency is bound by local process overhead (~50-200ms per invocation)","No built-in retry logic or timeout management — failures in the CLI process propagate directly to MCP client","Single-threaded CLI execution means concurrent requests queue sequentially, limiting throughput for high-volume code generation"],"requires":["Node.js 16+ (MCP server runtime)","OpenAI Codex CLI installed locally and in system PATH","Valid OpenAI API key (OPENAI_API_KEY environment variable)","MCP client compatible with stdio transport (Claude Desktop, custom MCP clients)"],"input_types":["text prompts (code generation requests)","code snippets (for context or completion)","CLI arguments passed through MCP call parameters"],"output_types":["generated code (text)","CLI stdout/stderr (text)","structured MCP response objects with result and metadata"],"categories":["tool-use-integration","code-generation-editing"],"confidence":0.5,"matches":0,"success_rate":0},{"id":"npm_npm-codex-mcp-server__cap_1","uri":"capability://tool.use.integration.mcp.resource.discovery.and.schema.advertisement","name":"mcp resource discovery and schema advertisement","description":"Implements the MCP server protocol to advertise Codex capabilities as discoverable resources with standardized schemas. The server registers itself with MCP clients, publishes available tools/resources with input/output schemas, and handles the MCP handshake protocol (initialization, capability negotiation) to enable clients like Claude to discover and invoke Codex without hardcoding tool definitions.","intents":["I want Claude or another MCP client to automatically discover Codex as an available tool without manual configuration","I need to define the input schema for Codex prompts so MCP clients can validate and structure requests correctly","I want to advertise Codex's capabilities (code generation, completion, etc.) in a standardized way across different MCP-compatible applications"],"best_for":["MCP client developers integrating multiple code generation tools and needing standardized discovery","teams deploying MCP servers in heterogeneous environments where tool schemas must be machine-readable","builders creating agent systems that dynamically load tools based on MCP resource advertisements"],"limitations":["Schema advertisement is static at server startup — no dynamic schema updates without server restart","MCP protocol overhead adds ~10-50ms per discovery/negotiation cycle compared to direct function calls","Limited to MCP-compatible clients; non-MCP tools cannot consume the advertised resources"],"requires":["MCP client library or compatible client (Claude Desktop, custom Node.js MCP client)","Understanding of MCP protocol (resources, tools, prompts concepts)","Node.js 16+ for the MCP server runtime"],"input_types":["MCP initialization messages","MCP resource/tool discovery requests"],"output_types":["MCP capability advertisements (JSON schemas)","resource/tool metadata with input/output specifications"],"categories":["tool-use-integration"],"confidence":0.5,"matches":0,"success_rate":0},{"id":"npm_npm-codex-mcp-server__cap_2","uri":"capability://tool.use.integration.environment.based.api.credential.injection.for.codex.cli","name":"environment-based api credential injection for codex cli","description":"Manages OpenAI API credentials by reading from environment variables (OPENAI_API_KEY) and injecting them into the Codex CLI process environment at invocation time. This approach avoids hardcoding secrets in configuration files and leverages Node.js process.env to pass credentials securely to child processes, with the MCP server acting as a credential broker between the client and the CLI.","intents":["I want to run Codex CLI with my OpenAI API key without storing it in code or config files","I need the MCP server to securely pass credentials to the Codex CLI subprocess without exposing them in logs or MCP messages","I want to rotate or update API keys without redeploying the MCP server"],"best_for":["developers deploying MCP servers in containerized or cloud environments where environment variables are the standard credential mechanism","teams using secret management systems (AWS Secrets Manager, HashiCorp Vault) that inject credentials as environment variables","solo developers running local MCP servers who want to avoid hardcoding API keys"],"limitations":["Credentials are passed as plaintext environment variables to child processes — vulnerable if process memory is inspected","No built-in credential rotation or expiration handling — relies on external secret management for key updates","Environment variable approach doesn't support per-request credential switching or multi-tenant scenarios","Credentials visible in process listings (ps aux) on Unix systems unless additional OS-level protections are in place"],"requires":["OPENAI_API_KEY environment variable set before MCP server startup","Valid OpenAI API key with Codex access","Node.js process.env access (standard in Node.js runtime)"],"input_types":["environment variables"],"output_types":["injected process environment for Codex CLI subprocess"],"categories":["tool-use-integration","safety-moderation"],"confidence":0.5,"matches":0,"success_rate":0},{"id":"npm_npm-codex-mcp-server__cap_3","uri":"capability://tool.use.integration.stdio.based.mcp.transport.and.bidirectional.message.routing","name":"stdio-based mcp transport and bidirectional message routing","description":"Implements the MCP server using stdio (standard input/output) as the transport layer, reading JSON-RPC messages from stdin and writing responses to stdout. This enables the MCP server to run as a subprocess of an MCP client (like Claude Desktop), with message routing handled by the MCP library's event loop that deserializes incoming requests, dispatches them to handler functions, and serializes responses back to the client.","intents":["I want to run the Codex MCP server as a subprocess of Claude Desktop without requiring a separate network service","I need bidirectional communication between the MCP client and server using a simple, process-based transport","I want to avoid network configuration (ports, firewalls, TLS) by using stdio for local inter-process communication"],"best_for":["Claude Desktop users running MCP servers locally without network infrastructure","developers building single-machine agent systems where stdio transport is sufficient","teams deploying MCP servers in containerized environments where stdio is the natural IPC mechanism"],"limitations":["Stdio transport is inherently single-client — only one MCP client can connect to a server instance at a time","No built-in multiplexing or connection pooling — concurrent requests from the same client are serialized by the MCP protocol","Debugging is harder than HTTP-based transports because stdio messages are not easily inspectable without logging","Process lifecycle is tightly coupled to the client — server crashes if the client terminates unexpectedly"],"requires":["MCP client that supports stdio transport (Claude Desktop, custom Node.js MCP clients)","Node.js 16+ with MCP server library (e.g., @modelcontextprotocol/sdk)","Proper stdio stream handling in the MCP server implementation"],"input_types":["JSON-RPC messages on stdin"],"output_types":["JSON-RPC responses on stdout"],"categories":["tool-use-integration"],"confidence":0.5,"matches":0,"success_rate":0},{"id":"npm_npm-codex-mcp-server__cap_4","uri":"capability://tool.use.integration.cli.argument.marshaling.and.parameter.passing.to.codex","name":"cli argument marshaling and parameter passing to codex","description":"Translates MCP request parameters (passed as JSON in the MCP call) into command-line arguments for the Codex CLI, handling parameter validation, type conversion, and argument formatting. The server constructs the appropriate CLI command string with flags and options based on the MCP request, then spawns the Codex process with these arguments, enabling MCP clients to control Codex behavior through structured parameter passing rather than raw CLI strings.","intents":["I want to pass code generation parameters (prompt, temperature, max tokens) to Codex through MCP without manually constructing CLI commands","I need the MCP server to validate and format parameters correctly before passing them to the Codex CLI","I want to expose Codex's CLI options (like model selection, output format) as discoverable MCP parameters"],"best_for":["MCP client developers who want to control Codex behavior through structured parameters rather than raw CLI strings","teams building code generation workflows where parameter validation and type safety are important","builders creating agent systems that need to dynamically adjust Codex settings based on task requirements"],"limitations":["Parameter support is limited to what the Codex CLI exposes — new CLI options require server code changes","No built-in parameter validation beyond basic type checking — complex validation logic must be implemented in the server","CLI argument escaping and quoting can be error-prone; special characters in parameters may cause CLI parsing issues","No parameter documentation or schema generation — clients must rely on hardcoded MCP resource schemas"],"requires":["Knowledge of Codex CLI argument format and available options","Node.js child_process module for spawning CLI with arguments","Codex CLI installed locally with known command-line interface"],"input_types":["JSON parameters from MCP request (strings, numbers, booleans)"],"output_types":["formatted CLI arguments passed to child process"],"categories":["tool-use-integration"],"confidence":0.5,"matches":0,"success_rate":0},{"id":"npm_npm-codex-mcp-server__cap_5","uri":"capability://tool.use.integration.subprocess.output.capture.and.response.marshaling","name":"subprocess output capture and response marshaling","description":"Captures stdout and stderr from the Codex CLI subprocess using Node.js stream handlers, buffers the output, and marshals it into MCP response objects with structured metadata (exit code, execution time, error status). The server handles both successful completions and error cases, converting raw CLI output into JSON-serializable MCP responses that can be transmitted back to the client with proper error handling and status codes.","intents":["I want the MCP server to capture Codex CLI output and return it as a structured response to the MCP client","I need to distinguish between successful code generation and CLI errors, and communicate both to the client","I want to include execution metadata (runtime, exit code) in the MCP response for debugging and monitoring"],"best_for":["MCP client developers who need structured responses from Codex with error handling and metadata","teams building observability into code generation workflows and needing execution details","builders creating agent systems that need to handle Codex failures gracefully"],"limitations":["Large output from Codex CLI may exceed buffer limits or cause memory issues if not streamed properly","Error messages from the CLI may be unstructured or unclear, making it hard to diagnose failures in the MCP client","No built-in retry logic — if the CLI fails, the error is immediately returned to the client without recovery attempts","Response marshaling adds ~10-50ms latency per invocation due to JSON serialization overhead"],"requires":["Node.js stream handling (readable streams, data/error event handlers)","Child process management (child_process.spawn or similar)","JSON serialization for MCP response objects"],"input_types":["stdout/stderr from Codex CLI subprocess"],"output_types":["MCP response objects with generated code, metadata, and error information"],"categories":["tool-use-integration"],"confidence":0.5,"matches":0,"success_rate":0}],"trust":{"score":27,"verified":false,"data_access_risk":"high","permissions":["Node.js 16+ (MCP server runtime)","OpenAI Codex CLI installed locally and in system PATH","Valid OpenAI API key (OPENAI_API_KEY environment variable)","MCP client compatible with stdio transport (Claude Desktop, custom MCP clients)","MCP client library or compatible client (Claude Desktop, custom Node.js MCP client)","Understanding of MCP protocol (resources, tools, prompts concepts)","Node.js 16+ for the MCP server runtime","OPENAI_API_KEY environment variable set before MCP server startup","Valid OpenAI API key with Codex access","Node.js process.env access (standard in Node.js runtime)"],"failure_modes":["Requires local Codex CLI installation and valid OpenAI API key — no fallback to cloud-only execution","Streaming output depends on Codex CLI's own streaming capabilities; latency is bound by local process overhead (~50-200ms per invocation)","No built-in retry logic or timeout management — failures in the CLI process propagate directly to MCP client","Single-threaded CLI execution means concurrent requests queue sequentially, limiting throughput for high-volume code generation","Schema advertisement is static at server startup — no dynamic schema updates without server restart","MCP protocol overhead adds ~10-50ms per discovery/negotiation cycle compared to direct function calls","Limited to MCP-compatible clients; non-MCP tools cannot consume the advertised resources","Credentials are passed as plaintext environment variables to child processes — vulnerable if process memory is inspected","No built-in credential rotation or expiration handling — relies on external secret management for key updates","Environment variable approach doesn't support per-request credential switching or multi-tenant scenarios","builder identity is not verified yet","no observed match outcomes yet"],"rank_breakdown":{"adoption":0.05,"quality":0.22,"ecosystem":0.48000000000000004,"match_graph":0.25,"freshness":0.6,"weights":{"adoption":0.25,"quality":0.25,"ecosystem":0.15,"match_graph":0.23,"freshness":0.12}},"observed_outcomes":{"matches":0,"success_rate":0,"avg_confidence":0,"top_intents":[],"last_matched_at":null},"maintenance":{"status":"active","updated_at":"2026-05-24T12:16:23.903Z","last_scraped_at":"2026-05-03T14:23:34.904Z","last_commit":null},"community":{"stars":null,"forks":null,"weekly_downloads":null,"model_downloads":null,"model_likes":null}},"distribution":{"claim_url":"https://unfragile.ai/submit?claim=npm-codex-mcp-server","compare_url":"https://unfragile.ai/compare?artifact=npm-codex-mcp-server"}},"signature":"ZKjPgGCZdFd9hwvSenLXm4hKRPHfB6oujCg+81CVeT9wP0Q77u3gEgv+siflTiPnmLf8RyK7RiITftx6I+coCw==","signedAt":"2026-06-22T08:40:33.850Z","signedBy":"unfragile.ai","version":1},"_links":{"self":"https://unfragile.ai/api/v1/passport/npm-codex-mcp-server","artifact":"https://unfragile.ai/npm-codex-mcp-server","verify":"https://unfragile.ai/api/v1/verify?slug=npm-codex-mcp-server","publicKey":"https://unfragile.ai/api/v1/trust-passport-public-key","spec":"https://unfragile.ai/trust","schema":"https://unfragile.ai/schema.json","docs":"https://unfragile.ai/docs"}}